14.

12 Recitation 2

September 28, 2012

Concepts
1. Rationality: formally, a player is said to be rational if and only if he maximizes
the expected value of his payoffs (given his beliefs about the other players’
strategies.)
2. Dominance: A strategy s∗i strictly dominates si if and only if

ui (si∗ , s−i ) > ui (si , s−i ), ∀s−i ∈ S−i .

3. Best response: For any player i, a strategy sBR i is a best response to s−i
if and only if
ui (sBR
i , s−i ) ≥ ui (si , s−i ), ∀si ∈ Si

This definition is identical to that of a dominant strategy except that it is

not for all s−i ∈ S−i but for a specific strategy s−i . If it were true for all s−i ,
then SiBR would also be a dominant strategy, which is a stronger requirement
than being a best response against some strategy s−i .
4. Nash Equilibrium: strategy profile (sN 1 , ...sN ) is a Nash Equilibrium if
E NE

and only if si is a best response to s−i = (s1 , ...sN

NE NE NE
i−1 , si+1 , ...sN ) for each
E NE NE

i. That is, for all i, we have that

ui (sN
i
E NE
, s−i ) ≥ ui (si , sN E
−i ) ∀si ∈ Si .

Problem 1 (Similar to HW1-4)

Suppose there is a (polluting) firm and a (pollution-averse) consumer. The firm
either pollutes or is shut down. One way for the (rich) government to resolve
the externality is as follows:
1. Ask the firm to state the monetary benefit b̂ of generating pollution
2. Ask the consumer to state the monetary equivalent of the cost of suffering
pollution, ĉ.
3. Shut the firm down iff ĉ ≥ b̂. If the firm is open, give the consumer b̂ and
charge the firm ĉ
The players are 1) the firm, and 2) the consumer. True benefit and cost are
b and c, respectively.

1
(a) Write this in the normal form.
The strategies
( are ˆb ∈ [0, ∞] and
( ĉ ∈ [0, ∞]. Utility
t ((pa
( yoffs) from strategy
ˆ ˆ ˆ
profile b, ĉ of players are uf b, ĉ = (b − ĉ) b > ĉ = (b − ĉ) 1[b̂>ĉ] and
( ( t (
uc ˆb, ĉ = ˆb − c ˆb > ĉ .

(b) Check if there is a dominant strategy equilibrium, and compute

it if there is one.
First, check the firm. Suppose b̂ > b. Then, three cases:
(
i) ˆb > b > ĉ: uf ˆb, ĉ = (b − ĉ) = uf (b, ĉ)
(
ii) ˆb > ĉ ≥ b: uf ˆb, ĉ = − (ĉ − b) < 0 = uf (b, ĉ)
(
iii) ĉ ≥ b̂ > b: uf ˆb, ĉ = 0 = uf (b, ĉ)
Now, suppose b > b̂
(
i) b > ˆb > ĉ: uf ˆb, ĉ = (b − ĉ) = uf (b, ĉ)
(
ii) b > ĉ ≥ b̂: uf ˆb, ĉ = 0 < b − ĉ = uf (b, ĉ)
(
iii) ĉ ≥ b > b̂: uf ˆb, ĉ = 0 = uf (b, ĉ)
( (
Hence, for any ˆb, uf (b, ĉ) ≥ uf ˆb, ĉ and there exist ĉ∗ ˆb such that
(
uf (b, ĉ∗ ) > uf ˆb, ĉ∗ .
Now, check the consumer.
( Suppose
( ĉ > c. Then,
i) ĉ > c ≥ ˆb : uc ˆb, ĉ = 0 = uc ˆb, c
( (
ii) ĉ ≥ ˆb > c : uc ˆb, ĉ = 0 < b̂ − c = uc ˆb, c
( (
iii) b̂ > ĉ > c : uc ˆb, ĉ = ˆb − c = uc ˆb, c
Finally, suppose ( c > ĉ : (
i) c > ĉ ≥ b : uc ˆb, ĉ = 0 = uc ˆb, c
ˆ
( ( (
ii) c ≥ ˆb > ĉ : uc ˆb, ĉ = − c − ˆb < 0 = uc ˆb, c
( (
iii) b̂ > c > cˆ : uc ˆb, ĉ = ˆb − c = uc ˆb, c
Therefore, both the consumer and the firm have truth-telling as a weakly
dominant strategy.
One problem: The government suffers a deficit of b̂ − ĉ, (or it may suffer a
surplus if the firm is closed if the government taxes everyone beforehand).

2
Problem 2 (2011 Midterm 1-1)
(a) Compute the set of all rationalizable strategies in the following
game.
w x y z
a 0,3 0,1 3,0 0,1
b 3,0 0,2 2,4 1,1
c 2,4 3,2 1,2 10,1
d 0,5 5,3 1,2 0,10

Answer : Iterated Elimination of Strictly Dominated Strategies: eliminate

all the strictly dominated strategies and iterate this k-times. In this procedure,
one eliminates all the strictly dominated strategies and iterates this k times.
Two main points are:
1. One must eliminate only the strictly dominated strategies. One cannot
eliminate a strategy if it is weakly dominated but not strictly dominated.
2. One must eliminate the strategies that are stricly dominated by mixed
strategies (but not necessarily by pure strategies).
Strategy x is strictly dominated by the mixed strategy σ2 with σ2 (w) ∈
1 1
and 2 (y) = 1 − σ2 (w). In the first round, x is therefore eliminated. (No


,
3 2 σ
other strategy is eliminated in that round.) In the second round, d is strictly
dominated by b and eliminated. In the third round, z is strictly dominated by
σ2 above and eliminated. In the fourth round, c is strictly dominated by b and
eliminated. There are no other elimination, and the set of rationalizable strate-
gies is {a,b}×{w,y}. (Note: explain how to find which strategy to eliminate by
checking best responses)

(b) Compute the set of all Nash equilibria.

Answer: The only Nash equilibrium is σ ∗ where σ1∗ (a) = 47 , σ1∗ (b) = 37 and
σ2∗ (w) = 14 , σ2∗ (y) = 34 . (Note: explain how to derive this - in order for one
player to mix strategies and play multiple strategies with positive probabilities,
he must be indifferent between those strategies, or have the same expected
utility for all choices)

Problem 3 (2009 HW2-1)

Consider the following investment game. There are two firms, each of them has
to decide how much to invest. If we let ki ≥ 0 be the investment choice of firm
i, then the profits of the firm are given by:

ki2
πi (ki , kj ) = Aki −
2
where the productivity of firm i is given by A = α + (ki + kj )(1 − α), where
α ∈ ( 23 , 1]. Note how productivity depends on the investment level of both
firms.

3
(a) What are the best response function for each firm as a function
of α?
Each firm will maximize its profit given the other firm’s strategy. The first order
condition is
∂πi
= α + kj (1 − α) + ki (1 − 2α) = 0
∂ki
In other words,
α + kj (1 − α)
ki = BRi (kj ) =
2α − 1

(b) Find the Nash equilibrium of the game, call it (k(α), k (α)). Note
that in the Nash equilibrium both firms choose the same investment
levels.
The Nash equilibrium is the point at which the two best response functions
intersect. By symmetry, we set ki = kj = k and solve

α + k(1 − α)
k=
2α − 1
(
The NE is α α
3α−2 , 3α−2 .

(c) What happens when α → 1? Does the equilibrium investment

level increase or decrease? Do you have intuition for this result?
2/3
α
k = 3α−2 = 13 + 3α−2 . Thus, as α → 1, the investment level decreases. Intuition
is clear: as α increases, productivity depends less on the total investment level
and firms want to invest less.

4
MIT OpenCourseWare
http://ocw.mit.edu

14.12 Economic Applications of Game Theory

Fall 2012

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.